At the core of autonomous vehicles are thousands of semiconductor chips that serve as the vehicle’s eyes, ears, and brain—sensing surroundings, making decisions, and controlling actions. As semiconductors are critical to the function and safety of autonomous vehicles, failure cannot be tolerated. All the chips need to work perfectly together without incident to protect the safety of both the vehicle’s occupants and others in the surrounding environment. This is the impetus for automobile manufacturers and Tier 1 suppliers requiring parts per billion levels of automotive semiconductor quality, and the force behind automotive IC manufacturers’ pursuit of Zero Defects standards.
Zero Defects is a comprehensive automotive quality initiative that extends to all car parts, including semiconductors. It enforces a quality mindset and continuous improvement. It includes a quality manufacturing process that is documented and can be audited. It includes requirements for robust design, thorough test, and a tightly monitored “safe launch” ramp up. It requires that all potential failure mechanisms be understood. If potential failures can’t be designed out, they must be controlled in the manufacturing process as described in a detailed inspection and metrology control plan.
Zero Defects manufacturing uses a specialized automotive process flow to ensure quality. This process flow requires more frequent inspection, tighter Statistical Process Control (SPC) limits, and extremely high process capability levels to ensure that all parts will meet the needed quality standards. Finally, any non-conforming material must be identified and excluded from the supply chain.
When applied to a semiconductor fab, Zero Defects has a profound impact on process control requirements and strategies. Small changes to established methods will not produce the IC quality necessary to be successful in the automotive IC market. As such, both IDMs and fabless/foundry manufacturing models are making fundamental changes to their historical approach to device yield and quality. By examining fundamental process control capability (e.g., critical defect types, inspection tool sensitivity), implementing more stringent defect reduction and excursion monitoring methods, and establishing new parts screening strategies, automotive fabs can make significant progress toward meeting Zero Defects goals. Several of the process control considerations and components for an automotive semiconductor fab are outlined below.
Latent Defects
One of the larger obstacles to Zero Defects success is the so-called “latent defect.” These are defects that impact chip reliability. Latent defects may be of a size or location that does not initially kill the die, or they may lie in an untested area of the die (an increasing problem with complex SoCs). As a result, the at-risk die passes electrical test and “escapes” into the supply chain. The demanding automotive environment of high heat, humidity, and vibration can sometimes “activate” these latent defects, causing a premature failure. The industry has long relied on electrical testing as the method to cull bad die, but latent defects pass electrical testing, so other methods are required to stop escapes near the source—in the fab—where costs are lower.
So, what are the characteristics of these latent defects and how do you find them? Failure analysis indicates that the majority of latent defects are, in fact, process-related defects that originate in the fab. In fact, the same defect types that impact yield (killer defects) also impact reliability—the two differ only in their size and/or where they occur on the device pattern. Because yield and reliability defects share the same root cause, establishing a comprehensive strategy in the fab to reduce random defects will both increase yield and improve reliability.
Inspection Sensitivity
When designing their process control strategy, automotive fabs need to decide on the minimum size of the defects that they want to detect and monitor—what is the smallest defect size that could cause future reliability issues? Even though yield and reliability defects share the same root cause, modelling results have shown that to control for, and reduce, the number of reliability defects present in the process, fabs need to capture smaller defects.
Therefore, automotive fabs need to go beyond the previous standards for reducing defectivity to optimize yield and implement higher sensitivity inspection strategies to optimize reliability. In general, detection of reliability defects requires an inspection sensitivity that is one node ahead of the current design node inspection plan for yield. Increasing the sensitivities of inspection recipes, or in some cases, using a more capable inspection system, will find smaller defects and possibly reveal previously hidden signatures of defectivity that represent an unacceptable risk to reliability for automotive fabs. With extra inspection sensitivity, automotive fabs can detect, monitor, and control the defects that would otherwise escape the fab and cause premature reliability failures.
Continuous Improvement Program for Baseline Defect Reduction
The best first approach to manufacturing devices with fewer overall defects is to closely control the process by employing continuous improvements programs that reduce the random defectivity introduced by the process tools or environment. This requires implementing fundamental baseline defect reduction techniques—primarily, tool monitoring. Tool monitoring is the established best practice for isolating the source of random defectivity contributed by the fab’s process tools. During tool monitoring, a bare wafer is inspected to establish its baseline defectivity, run through a specific process tool (or chamber), and then inspected again. Any defects that were added to the wafer must have come from that specific process tool. This method can reveal the cleanest “golden” tools in the fab, as well as the “dog” tools that contribute the most defects and require corrective action. With plots of historical defect data from the process tools, goals and milestones for continuous improvement in defect reduction can be implemented.
Fabs have used tool monitoring strategies for years, but automotive fabs need to raise the technique to a higher standard to achieve the lower defect levels necessary to improve IC reliability. Only by implementing a methodical defect reduction program will a fab move toward the
Zero Defects goal and be able to pass the stringent audits required by automobile manufacturers.
Sampling and Traceability
Following baseline defect reduction, the next best approach to control reliability defects is implementing an adequate sample plan. The sample plan must be set for the right process steps at sufficient frequency to quickly flag process or tool excursions. Additionally, there should be sufficient inspection capacity to support a control plan that expedites excursion detection, root cause isolation, and WIP-at-risk traceability. To meet these objectives, the control plan of an automotive process flow will invariably require more inspection steps and more frequent sampling (both as a percentage of lots and number of wafers per lot) than the control plan for production of ICs for consumer products. With higher sampling when the inevitable process excursion happens, the automotive fab will know definitively where the problem started and stopped. Thus, it can quarantine the affected parts until they can be effectively dispositioned or culled—thereby ensuring that non-conforming devices will not inadvertently ship.
Inline Defect Part Average Testing
A method that is receiving increasing interest in the automotive industry is the utilization of inline defect information not only to control the process, but also to identify die at risk for reliability problems while they are still in the fab, where the cost of correcting the problem is the lowest. Automotive fabs have long relied on “screening”—where a high throughput tool inspects 100 percent of the die on all wafers at a handful of final layers late in the manufacturing process. Die that meet the defined failure criteria (defect size/type/location) are excluded or “inked.” In the past, this method alone could result in higher than desired levels of “overkill,” making it an expensive choice for quality enhancement.
A new inline technique, called I-PAT™ (Inline Defect Part Average Testing), may be the answer. It leverages a 20-year-old automotive industry technique known as Part Average Testing. This original method, based on e-test, identifies any die whose test results lie outside of the normal distribution of the population, even if they are within the operating specifications. For a small sacrifice of 0.5 to 2.5 percent yield, significant improvements in reliability are gained, with some seeing 20 to 30 percent improvement when these outlier die are culled. I-PAT™ moves this concept inline, looking for die with outlier defect populations across the stacked critical screening inspection steps. These outlier die are statistically more likely to contain the latent defects that the industry desperately wants to eliminate. I-PAT results could be used to cull these at-risk die. Or, I-PAT™ results can be combined later with electrical outlier methods to improve the overall go/no-go decision for die.
Process control plays a key role in helping automotive semiconductor manufacturers achieve Zero Defects success. By implementing appropriate inspection tools, robust sampling strategies, and continuous improvement programs, fabs have a higher probability of finding and reducing the random defectivity that leads to reliability failures in automotive ICs. Innovative inline screening techniques further prevent at-risk devices from entering the supply stream. These comprehensive control strategies help ensure that the thousands of chips behind the wheel of autonomous vehicles are stable and safe.
